24_deb_pkg_gov/new_contrib_window.py

import csv
from perceval.backends.core.git import Git
import os
import datetime as dt
import time
import shutil 
import pandas as pd
import dateutil
from tqdm import tqdm
import math

key = os.environ.get('KKEXKEY')

early_cutoff = dt.datetime(2008,2, 8)
temp_dir = "/data/users/mgaughan/tmp/"

'''
- rate of change, rate of all/day
'''
def file_get_contribs(upstream_vcs_link, event_date):
    # if we're looking at readmes me_read is true and if not, if we're looking at contributing files, it's false
    #this is the window of days on either side of the event that we're looking at
    window = 56
    #print(upstream_vcs_link.split('/')[4])
    project_dict = {}
    project_dict['upstream_vcs_link'] = upstream_vcs_link
    project_dict['event_date'] = event_date
    upstream_vcs_link = upstream_vcs_link.strip()
    if "github" in upstream_vcs_link or "gitlab" in upstream_vcs_link:
        #making an evaluation that sub branches aren't being used and that people would fork if needed
        #this only looks at main
        upstream_vcs_link = "/".join(upstream_vcs_link.split("/")[0:5])
        print(upstream_vcs_link)
        full_temp_path = temp_dir + upstream_vcs_link.split('/')[4] + ".git"
    else:
        full_temp_path = temp_dir + upstream_vcs_link.split('/')[- 1] + ".git"
    print(upstream_vcs_link)
    if upstream_vcs_link == "https://gitlab.com/ubports/core" or upstream_vcs_link == "https://gitlab.freedesktop.org/xorg/lib":
        shutil.rmtree(full_temp_path, ignore_errors=True)
        return {}
    repo = Git(uri=upstream_vcs_link, gitpath=full_temp_path)
    try:
        commits = repo.fetch()
    except:
        print("perceval issue")
        return
    merge_pre, merge_post = 0, 0
    #list of tuples which has date and whether it was a merge
    commit_list = []
    first_date_readme = ""
    for commit in commits:
        #print(commit['data'])
        if "Merge" in commit['data'].keys():
            commit_list.append([commit['data']['CommitDate'], True, commit['data']['Author'], commit['data']['Commit']])
            if dateutil.parser.parse(commit['data']['CommitDate']) < event_date:
                merge_post += 1
            else:
                merge_pre += 1
        else: 
            commit_list.append([commit['data']['CommitDate'], False, commit['data']['Author'], commit['data']['Commit']])
    shutil.rmtree(full_temp_path, ignore_errors=True)
    project_dict['merge_pre'] = merge_pre
    project_dict['merge_post'] = merge_post
    collab_roster = []
    contrib_roster = []
    new_collab_pre = 0
    new_collab_post = 0
    new_contrib_pre = 0
    new_contrib_post = 0
    for commit in commit_list:
        commit_td = dateutil.parser.parse(commit[0])
        if commit_td >= (event_date + dt.timedelta(days=-window, hours=0)) and commit_td <= event_date:
            if commit[3] not in collab_roster:
                new_collab_pre += 1
            if commit[2] not in contrib_roster:
                new_contrib_pre += 1
        if commit_td > event_date and commit_td <= (event_date + dt.timedelta(days=window, hours=0)):
            if commit[3] not in collab_roster:
                new_collab_post += 1
            if commit[2] not in contrib_roster:
                new_contrib_post += 1
        if commit[3] not in collab_roster:
            collab_roster.append(commit[3])
        if commit[2] not in contrib_roster:
            contrib_roster.append(commit[2])
    project_dict["before_contrib_new"] = new_contrib_pre
    project_dict['after_contrib_new'] = new_contrib_post
    project_dict['before_collab_new'] = new_collab_pre
    project_dict['after_collab_new'] = new_collab_post
    print(project_dict)
    return project_dict



def for_files():
    csv_path = "final_data/deb_readme_did.csv"
    count = 0 
    with open(csv_path, 'r') as file:
        csv_reader = csv.DictReader(file)
        with open('42524_new_contrib.csv', "w") as writing_file:
            # this would also have to get switched fro the cont dataset
            keys = ['upstream_vcs_link', "event_date", 'before_contrib_new', 'after_contrib_new', 'after_collab_new', 'before_collab_new', 'merge_pre', 'merge_post']
            dict_writer = csv.DictWriter(writing_file, keys)
            dict_writer.writeheader()
            for row in csv_reader:
                count += 1
                print(row['upstream_vcs_link'])
                # this would have to get switched to false for the cont dataset
                try:
                    dict_row = file_get_contribs(row['upstream_vcs_link'].strip(), row['event_date'])
                except:
                    dict_row = {}
                dict_writer.writerow(dict_row)
    

if __name__ == "__main__":
    for_files()
    #file_get_pr("https://github.com/tqdm/tqdm", True)
    #file_get_pr("https://github.com/GameServerManagers/LinuxGSM", True)
    #file_get_pr("https://github.com/walling/unorm/issues/new/", True)
    #file_get_pr("https://github.com/krahets/hello-algo/tree/dev1", True)